Comprehensive BPF offload
|
|
- Homer Jackson
- 5 years ago
- Views:
Transcription
1 Comprehensive BPF offload Nic Viljoen & Jakub Kicinski Netronome Systems An NFP based NIC (1U) netdev 2.2, Seoul, South Korea November 8-10, 2017
2 Agenda Refresher Programming Model Architecture Performance & Optimization Requirements for Production Offload bpftool Verifier restructuring
3 Programming Model (refresher) Program is written in standard manner LLVM compiled as normal iproute/tc loads the program requesting offload The nfp_bpf_jit.c converts the ebpf bytecode to NFP machine code (and we mean the actual machine code :)) Translation reuses a significant amount of verifier infrastructure
4 Refresher-BPF Offload Mapping GPRs 10 Registers (64 bit, 32 bit subregisters) LMEM (1 KB) Thread (x4 per Core) CLS (64 KB) Driver Core (x60 used for BPF) 512 byte stack CTM (256 KB) Island (x6 per Chip) Maps, varying sizes IMEM(4 MB) Chip DRAM (2GB) NIC
5 Performance Simple XDP load balancer (~ 800 BPF insns, 4 lookups) Based on the TC example in kernel - selftests/bpf/l4lb.c Combined with samples/bpf/xdp_tx_iptunnel_kern.c Per CPU array changed to standard array to run offloaded There is no nice equivalent for per CPU at the moment on the NIC Not optimised-big health warning :)
6 Future Optimizations Map placement/caching-as shown on previous page Using Packet Cache-reduce latency of packet accesses from ~50 cycles to ~3 cycles 32 bit ALU from LLVM where possible-reduce ALUs from ~ 4 machine code insns to 1 Remove FW locks-double memory bandwidth
7 Requirements For Production Readiness Multi-stage processing: Reliable manner to run some programs in host if not possible/desirable in offload Debug: Usable verifier error messages Introspection-both of maps and programs JIT: Translation before optimization
8 Multi-Stage Processing Offload some programs This can be managed by the driver (implicitly) or explicitly Use data_meta to inject programs into the correct BPF program to run next Important for edge case where the next program to run is not fixed Allows offload to be used for beneficial cases only Can be explicitly via flags 2017 NETRONOME SYSTEMS, INC. CONFIDENTIAL 8
9 Progress made - kernel Upstream: new instructions (Daniel, Jiong, I); direct packet access; stack support; adjust head helper; add 32-bit subregister support to LLVM (Jiong). Prototyped/PoC: map offload support (hash and array maps); atomic add operation; memcpy optimizations (Jiong); initiate work on register state tracking (Jiong).
10 Progress made - tooling bpftool in kernel tree for Linux 4.15 (in the tools/ directory); iproute2-like syntax; list and pin objects; programs: show type, name, tag, id, memory usage, load time, used maps; dump JITed and translated images (to file or print instructions); maps: show type, name, id, key/value size, number of elements, flags; lookup, update, delete, etc. JSON output (Quentin); BPF FS integration (Quentin, Prashant).
11 Progress made - tooling llvm-mc (Jiong) upstream for LLVM 6.0; LLVM s macro assembler; verifier-style syntax: r1 = r6 r2 = 0xff ll call 12 r0 = 0 exit allows hand-crafting precise BPF programs (or compiling C code into assembly and modifying it); opens way for BPF inline assembly; very useful for testing particular instruction sequences.
12 Kernel basics (refresher) user space kernel space BPF syscall program type (sk filter, kprobe, cls, xdp) license... fd fd, skip_* flags tc fd, skip_* flags XDP ctrl verifier BPF prog TC cls_bpf verification host JIT offload object modification HW JIT / translator ndo setup tc stats & maps RX XDP TX driver
13 Translation and loading (refresher) user space kernel space BPF syscall program type (sk filter, kprobe, cls, xdp) license... fd fd, skip_* flags tc fd, skip_* flags XDP ctrl verifier BPF prog (3) Collect state/analyze TC cls_bpf verification modification host JIT offload object (4) Optimize (5) JIT/generate image (6) Load image (2) Re-run the verifier HW JIT / translator ndo setup tc stats & maps RX XDP TX driver (1) Check HW capabilities and image parameters
14 Kernel basics (refresher) user space kernel space BPF syscall program type (sk filter, kprobe, cls, xdp) license... fd fd, skip_* flags tc fd, skip_* flags XDP ctrl verifier BPF prog TC cls_bpf verification modification host JIT offload object HW JIT / translator ndo setup tc stats & maps RX XDP TX driver
15 Progress made - kernel user space kernel space BPF syscall program type (sk filter, kprobe, cls, xdp) license ifindex... fd XDP ctrl tc driver ops find device XDP TC verifier BPF prog offload verifier prep translate offload object cls_bpf verification destroy modification JIT RX XDP TX ndo setup tc driver driver
16 Rationale for recent kernel changes allow device translator to access the loaded program as-is: IDs/offsets not translated: structure field offsets; functions; map IDs. no prolog/epilogue injected; no optimizations made; output errors at program load and map creation time; make use of access to the verifier log; include device information in introspection APIs (bpftool); dump translated image: similar to host JITed image ; BPF core already has access to offload state (no longer driver black box); need to report machine info?
17 Debug and tooling APIs netlink extack support in cls_bpf/tc offloads: XDP already carries extack for use by the drivers; allows easier error reporting at attachment time; bpf perf event output - output samples for debugging the datapath; simple API for enabling/disabling optimizations: verifier/kernel already has some simple optimizations (e.g. lookup inlining); nfp translator already has a few and we expect to add more; need to report, enable/disable optimizations with nice granularity; maps: create maps on the device from the start; simplify map load/eviction and locking greatly; report errors/resource exhaustion at map creation time.
18 The end Thank you!
The Challenges of XDP Hardware Offload
FOSDEM 18 Brussels, 2018-02-03 The Challenges of XDP Hardware Offload Quentin Monnet @qeole ebpf and XDP Q. Monnet XDP Hardware Offload 2/29 ebpf, extended Berkeley Packet
More informationebpf Offload to Hardware cls_bpf and XDP
ebpf Offload to Hardware cls_bpf and Nic Viljoen, DXDD (Based on Netdev 1.2 talk) November 10th 2016 1 What is ebpf? A universal in-kernel virtual machine 10 64-bit registers 512 byte stack Infinite size
More informationXDP Hardware Offload: Current Work, Debugging and Edge Cases
XDP Hardware Offload: Current Work, Debugging and Edge Cases Jakub Kicinski, Nicolaas Viljoen Netronome Systems Santa Clara, United States jakub.kicinski@netronome.com, nick.viljoen@netronome.com Abstract
More informationBPF Hardware Offload Deep Dive
BPF Hardware Offload Deep Dive Jakub Kicinski 2018 NETRONOME SYSTEMS, INC. BPF Sandbox As a goal of BPF IR JITing of BPF IR to most RISC cores should be very easy BPF VM provides a simple and well understood
More informationSmartNIC Data Plane Acceleration & Reconfiguration. Nic Viljoen, Senior Software Engineer, Netronome
SmartNIC Data Plane Acceleration & Reconfiguration Nic Viljoen, Senior Software Engineer, Netronome Need for Accelerators Validated by Mega-Scale Operators Large R&D budgets, deep acceleration software
More informationHost Dataplane Acceleration: SmartNIC Deployment Models
Host Dataplane Acceleration: SmartNIC Deployment Models Simon Horman 20 August 2018 2018 NETRONOME SYSTEMS, INC. Agenda Introduction Hardware and Software Switching SDN Programmability Host Datapath Acceleration
More informationebpf Tooling and Debugging Infrastructure
ebpf Tooling and Debugging Infrastructure Quentin Monnet Fall ebpf Webinar Series 2018-10-09 Netronome 2018 Injecting Programs into the Kernel ebpf programs are usually compiled from C (or Go, Rust, Lua
More informationXDP For the Rest of Us
이해력 XDP For the Rest of Us Jesper Dangaard Brouer - Principal Engineer, Red Hat Andy Gospodarek - Principal Engineer, Broadcom Netdev 2.2, November 8th, 2017 South Korea, Seoul Motivation for this talk
More informationebpf Debugging Infrastructure Current Techniques and Additional Proposals
BPF Microconference 2018-11-15 ebpf Debugging Infrastructure Current Techniques and Additional Proposals Quentin Monnet Debugging Infrastructure What do we want to debug,
More informationAccelerating VM networking through XDP. Jason Wang Red Hat
Accelerating VM networking through XDP Jason Wang Red Hat Agenda Kernel VS userspace Introduction to XDP XDP for VM Use cases Benchmark and TODO Q&A Kernel Networking datapath TAP A driver to transmit
More informationBringing the Power of ebpf to Open vswitch. Linux Plumber 2018 William Tu, Joe Stringer, Yifeng Sun, Yi-Hung Wei VMware Inc. and Cilium.
Bringing the Power of ebpf to Open vswitch Linux Plumber 2018 William Tu, Joe Stringer, Yifeng Sun, Yi-Hung Wei VMware Inc. and Cilium.io 1 Outline Introduction and Motivation OVS-eBPF Project OVS-AF_XDP
More informationebpf Offload Getting Started Guide
ebpf Offload Getting Started Guide Netronome CX SmartNIC Revision 1.0 April 2018 ebpf Offload 1 Introduction 3 Kernel version support 4 Environment Setup 5 Kernel 5 Firmware 6 Driver 7 Setting up rings
More informationAccelerating Load Balancing programs using HW- Based Hints in XDP
Accelerating Load Balancing programs using HW- Based Hints in XDP PJ Waskiewicz, Network Software Engineer Neerav Parikh, Software Architect Intel Corp. Agenda Overview express Data path (XDP) Software
More informationA practical introduction to XDP
A practical introduction to XDP Jesper Dangaard Brouer (Red Hat) Andy Gospodarek (Broadcom) Linux Plumbers Conference (LPC) Vancouver, Nov 2018 1 What will you learn? Introduction to XDP and relationship
More informationOpen-NFP Summer Webinar Series: Session 4: P4, EBPF And Linux TC Offload
Open-NFP Summer Webinar Series: Session 4: P4, EBPF And Linux TC Offload Dinan Gunawardena & Jakub Kicinski - Netronome August 24, 2016 1 Open-NFP www.open-nfp.org Support and grow reusable research in
More informationBringing the Power of ebpf to Open vswitch
Bringing the Power of ebpf to Open vswitch William Tu 1 Joe Stringer 2 Yifeng Sun 1 Yi-Hung Wei 1 u9012063@gmail.com joe@cilium.io pkusunyifeng@gmail.com yihung.wei@gmail.com 1 VMware Inc. 2 Cilium.io
More informationDPDK+eBPF KONSTANTIN ANANYEV INTEL
x DPDK+eBPF KONSTANTIN ANANYEV INTEL BPF overview BPF (Berkeley Packet Filter) is a VM in the kernel (linux/freebsd/etc.) allowing to execute bytecode at various hook points in a safe manner. It is used
More informationXDP: 1.5 years in production. Evolution and lessons learned. Nikita V. Shirokov
XDP: 1.5 years in production. Evolution and lessons learned. Nikita V. Shirokov Facebook Traffic team Goals of this talk: Show how bpf infrastructure (maps/helpers) could be used for building networking
More informationENG. WORKSHOP: PRES. Linux Networking Greatness (part II). Roopa Prabhu/Director Engineering Linux Software/Cumulus Networks.
ENG. WORKSHOP: PRES. Linux Networking Greatness (part II). Roopa Prabhu/Director Engineering Linux Software/Cumulus Networks. 3 4 Disaggregation Native Linux networking (= server/host networking) Linux
More informationXDP: The Future of Networks. David S. Miller, Red Hat Inc., Seoul 2017
XDP: The Future of Networks David S. Miller, Red Hat Inc., Seoul 2017 Overview History of ebpf and XDP Why is it important. Fake News about ebpf and XDP Ongoing improvements and future developments Workflow
More informationSmartNIC Programming Models
SmartNIC Programming Models Johann Tönsing 206--09 206 Open-NFP Agenda SmartNIC hardware Pre-programmed vs. custom (C and/or P4) firmware Programming models / offload models Switching on NIC, with SR-IOV
More informationXDP now with REDIRECT
XDP - express Data Path XDP now with REDIRECT Jesper Dangaard Brouer, Principal Engineer, Red Hat Inc. LLC - Lund Linux Conf Sweden, Lund, May 2018 Intro: What is XDP? Really, don't everybody know what
More informationQuick Introduction to ebpf & BCC Suchakrapani Sharma. Kernel Meetup (Reserved Bit, Pune)
Quick Introduction to ebpf & BCC Suchakrapani Sharma Kernel Meetup (Reserved Bit, Pune) 24 th June 2017 ebpf Stateful, programmable, in-kernel decisions for networking, tracing and security One Ring by
More informationSmartNIC Programming Models
SmartNIC Programming Models Johann Tönsing 207-06-07 207 Open-NFP Agenda SmartNIC hardware Pre-programmed vs. custom (C and/or P4) firmware Programming models / offload models Switching on NIC, with SR-IOV
More informationOn getting tc classifier fully programmable with cls bpf.
On getting tc classifier fully programmable with cls bpf. Daniel Borkmann Noiro Networks / Cisco netdev 1.1, Sevilla, February 12, 2016 Daniel Borkmann tc and cls bpf with ebpf February
More informationProgramming NFP with P4 and C
WHITE PAPER Programming NFP with P4 and C THE NFP FAMILY OF FLOW PROCESSORS ARE SOPHISTICATED PROCESSORS SPECIALIZED TOWARDS HIGH-PERFORMANCE FLOW PROCESSING. CONTENTS INTRODUCTION...1 PROGRAMMING THE
More informationNetronome NFP: Theory of Operation
WHITE PAPER Netronome NFP: Theory of Operation TO ACHIEVE PERFORMANCE GOALS, A MULTI-CORE PROCESSOR NEEDS AN EFFICIENT DATA MOVEMENT ARCHITECTURE. CONTENTS 1. INTRODUCTION...1 2. ARCHITECTURE OVERVIEW...2
More informationProgramming Netronome Agilio SmartNICs
WHITE PAPER Programming Netronome Agilio SmartNICs NFP-4000 AND NFP-6000 FAMILY: SUPPORTED PROGRAMMING MODELS THE AGILIO SMARTNICS DELIVER HIGH- PERFORMANCE SERVER- BASED NETWORKING APPLICATIONS SUCH AS
More informationCombining ktls and BPF for Introspection and Policy Enforcement
Combining ktls and BPF for Introspection and Policy Enforcement Daniel Borkmann, John Fastabend Cilium.io Linux Plumbers 2018, Vancouver, Nov 14, 2018 Daniel Borkmann, John Fastabend ktls and BPF Nov 14,
More informationOVS Acceleration using Network Flow Processors
Acceleration using Network Processors Johann Tönsing 2014-11-18 1 Agenda Background: on Network Processors Network device types => features required => acceleration concerns Acceleration Options (or )
More informationComparing Open vswitch (OpenFlow) and P4 Dataplanes for Agilio SmartNICs
Comparing Open vswitch (OpenFlow) and P4 Dataplanes for Agilio SmartNICs Johann Tönsing May 24, 206 206 NETRONOME Agenda Contributions of OpenFlow, Open vswitch and P4 OpenFlow features missing in P4,
More informationXDP: a new fast and programmable network layer
XDP - express Data Path XDP: a new fast and programmable network layer Jesper Dangaard Brouer Principal Engineer, Red Hat Inc. Kernel Recipes Paris, France, Sep 2018 What will you learn? What will you
More informationSuricata Performance with a S like Security
Suricata Performance with a S like Security É. Leblond Stamus Networks July. 03, 2018 É. Leblond (Stamus Networks) Suricata Performance with a S like Security July. 03, 2018 1 / 31 1 Introduction Features
More informationVirtual Switch Acceleration with OVS-TC
WHITE PAPER Virtual Switch Acceleration with OVS-TC HARDWARE ACCELERATED OVS-TC PROVIDES BETTER CPU EFFICIENCY, LOWER COMPLEXITY, ENHANCED SCALABILITY AND INCREASED NETWORK PERFORMANCE COMPARED TO KERNEL-
More informationIntroducing the Cray XMT. Petr Konecny May 4 th 2007
Introducing the Cray XMT Petr Konecny May 4 th 2007 Agenda Origins of the Cray XMT Cray XMT system architecture Cray XT infrastructure Cray Threadstorm processor Shared memory programming model Benefits/drawbacks/solutions
More informationOn getting tc classifier fully programmable with cls bpf.
On getting tc classifier fully programmable with cls bpf. Daniel Borkmann Cisco Switzerland daniel@iogearbox.net Abstract Berkely Packet Filter, short BPF, is an instruction set architecture that was designed
More informationAgilio CX 2x40GbE with OVS-TC
PERFORMANCE REPORT Agilio CX 2x4GbE with OVS-TC OVS-TC WITH AN AGILIO CX SMARTNIC CAN IMPROVE A SIMPLE L2 FORWARDING USE CASE AT LEAST 2X. WHEN SCALED TO REAL LIFE USE CASES WITH COMPLEX RULES TUNNELING
More informationNetronome 25GbE SmartNICs with Open vswitch Hardware Offload Drive Unmatched Cloud and Data Center Infrastructure Performance
WHITE PAPER Netronome 25GbE SmartNICs with Open vswitch Hardware Offload Drive Unmatched Cloud and NETRONOME AGILIO CX 25GBE SMARTNICS SIGNIFICANTLY OUTPERFORM MELLANOX CONNECTX-5 25GBE NICS UNDER HIGH-STRESS
More informationDesign, Verification and Emulation of an Island-Based Network Flow Processor
Design, Verification and Emulation of an Island-Based Network Flow Processor Ron Swartzentruber CDN Live April 5, 2016 1 2016 NETRONOME SYSTEMS, INC. Problem Statements 1) Design a large-scale 200Gbps
More informationTracing Lustre. New approach to debugging. ORNL is managed by UT-Battelle for the US Department of Energy
Tracing Lustre New approach to debugging ORNL is managed by UT-Battelle for the US Department of Energy Current Lustre debugging tools Utility lctl handles profiling developed long before standard kernel
More informationFast packet processing in linux with af_xdp
Fast packet processing in linux with af_xdp Magnus Karlsson and Björn Töpel, Intel Legal Disclaimer Intel technologies may require enabled hardware, specific software, or services activation. Check with
More informationUsing SR-IOV offloads with Open-vSwitch and similar applications
Using SR-IOV offloads with Open-vSwitch and similar applications Rony Efraim, Or Gerlitz Netdev conferences 1.2 Oct 2016 Agenda Solution building blocks Use the software datapath (kernel) as a slow path
More informationCaches and Memory Hierarchy: Review. UCSB CS240A, Winter 2016
Caches and Memory Hierarchy: Review UCSB CS240A, Winter 2016 1 Motivation Most applications in a single processor runs at only 10-20% of the processor peak Most of the single processor performance loss
More informationrtnl mutex, the network stack big kernel lock
rtnl mutex, the network stack big kernel lock Red Hat Florian Westphal 4096R/AD5FF600 fw@strlen.de 80A9 20C5 B203 E069 F586 AE9F 7091 A8D9 AD5F F600 netdev 2.2, Seoul, November 2017 Agenda 1 Intro: What
More informationVirtual to physical address translation
Virtual to physical address translation Virtual memory with paging Page table per process Page table entry includes present bit frame number modify bit flags for protection and sharing. Page tables can
More informationXen Network I/O Performance Analysis and Opportunities for Improvement
Xen Network I/O Performance Analysis and Opportunities for Improvement J. Renato Santos G. (John) Janakiraman Yoshio Turner HP Labs Xen Summit April 17-18, 27 23 Hewlett-Packard Development Company, L.P.
More informationLinux Network Programming with P4. Linux Plumbers 2018 Fabian Ruffy, William Tu, Mihai Budiu VMware Inc. and University of British Columbia
Linux Network Programming with P4 Linux Plumbers 2018 Fabian Ruffy, William Tu, Mihai Budiu VMware Inc. and University of British Columbia Outline Introduction to P4 XDP and the P4 Compiler Testing Example
More informationEfficient JIT to 32-bit Arches
Efficient JIT to 32-bit Arches Jiong Wang Linux Plumbers Conference Vancouver, Nov, 2018 1 Background ISA specification and impact on JIT compiler Default code-gen use 64-bit register, ALU64, JMP64 test_l4lb_noinline.c
More informationOpen vswitch: A Whirlwind Tour. Jus8n Pe:t March 3, 2011
Open vswitch: A Whirlwind Tour Jus8n Pe:t March 3, 2011 Overview Visibility (NetFlow, sflow, SPAN/RSPAN) Fine- grained ACLs and QoS policies Centralized control through OpenFlow Port bonding, LACP, tunneling
More informationCaches and Memory Hierarchy: Review. UCSB CS240A, Fall 2017
Caches and Memory Hierarchy: Review UCSB CS24A, Fall 27 Motivation Most applications in a single processor runs at only - 2% of the processor peak Most of the single processor performance loss is in the
More informationHigh Performance Computing on GPUs using NVIDIA CUDA
High Performance Computing on GPUs using NVIDIA CUDA Slides include some material from GPGPU tutorial at SIGGRAPH2007: http://www.gpgpu.org/s2007 1 Outline Motivation Stream programming Simplified HW and
More informationThe Path to DPDK Speeds for AF XDP
The Path to DPDK Speeds for AF XDP Magnus Karlsson, magnus.karlsson@intel.com Björn Töpel, bjorn.topel@intel.com Linux Plumbers Conference, Vancouver, 2018 Legal Disclaimer Intel technologies may require
More informationAn Implementation Of Multiprocessor Linux
An Implementation Of Multiprocessor Linux This document describes the implementation of a simple SMP Linux kernel extension and how to use this to develop SMP Linux kernels for architectures other than
More informationCombining ktls and BPF for Introspection and Policy Enforcement
Combining ktls and BPF for Introspection and Policy Enforcement ABSTRACT Daniel Borkmann Cilium.io daniel@cilium.io Kernel TLS is a mechanism introduced in Linux kernel 4.13 to allow the datapath of a
More informationebpf & Switch Abstractions
Networking Track Vancouver, 14 November 2018 ebpf & Switch Abstractions Nick Viljoen "2018"NETRONOME"SYSTEMS,"INC. 1 Contents Background The"MultiCHost"NIC"Abstraction The"Switchdev"Based"MultiCHost"NIC"Abstraction
More informationAgilio OVS Software Architecture
WHITE PAPER Agilio OVS Software Architecture FOR SERVER-BASED NETWORKING THERE IS CONSTANT PRESSURE TO IMPROVE SERVER- BASED NETWORKING PERFORMANCE DUE TO THE INCREASED USE OF SERVER AND NETWORK VIRTUALIZATION
More informationOpenContrail, Real Speed: Offloading vrouter
OpenContrail, Real Speed: Offloading vrouter Chris Telfer, Distinguished Engineer, Netronome Ted Drapas, Sr Director Software Engineering, Netronome 1 Agenda Introduction to OpenContrail & OpenContrail
More information#ebpf You Cannot Stop This
#ebpf You Cannot Stop This David S. Miller, Red Hat Inc. davem_dokebi Van, The Man Ultimately, the true father of container networking and security Invented BPF in 1992 with Steven McCanne at LBL Recognized
More informationTraceLeft. A Configuration Driven ebpf Tracing Framework. Suchakra Sharma & Alban Crequy All Systems Go, 29th September 2018, Berlin
TraceLeft A Configuration Driven ebpf Tracing Framework Suchakra Sharma & Alban Crequy All Systems Go, 29th September 2018, Berlin Suchakra Sharma Alban Crequy Github: tuxology Twitter: @tuxology Email:
More informationHigh Performance Computing Lecture 26. Matthew Jacob Indian Institute of Science
High Performance Computing Lecture 26 Matthew Jacob Indian Institute of Science Agenda 1. Program execution: Compilation, Object files, Function call and return, Address space, Data & its representation
More informationERSPAN in Linux. A short history and review. Presenters: William Tu and Greg Rose
ERSPAN in Linux A short history and review. Presenters: William Tu and Greg Rose 1 What is Port Mirroring? Port mirroring is one of the most common network troubleshooting techniques. SPAN - Switch Port
More informationAn Introduction to the QorIQ Data Path Acceleration Architecture (DPAA) AN129
July 14, 2009 An Introduction to the QorIQ Data Path Acceleration Architecture (DPAA) AN129 David Lapp Senior System Architect What is the Datapath Acceleration Architecture (DPAA)? The QorIQ DPAA is a
More informationRALPH BÖHME, SERNET, SAMBA TEAM UNDERSTANDING AND IMPROVING SAMBA FILESERVER PERFORMANCE HOW I FELL IN LOVE WITH SYSTEMTAP AND PERF
UNDERSTANDING AND IMPROVING HOW I FELL IN LOVE WITH SYSTEMTAP AND PERF 2 AGENDA Disclaimer: focus on userspace, not kernel, mostly Linux Linux tracing history tour de force perf Systemtap Samba fileserver
More informationLinux Network Tuning Guide for AMD EPYC Processor Based Servers
Linux Network Tuning Guide for AMD EPYC Processor Application Note Publication # 56224 Revision: 1.00 Issue Date: November 2017 Advanced Micro Devices 2017 Advanced Micro Devices, Inc. All rights reserved.
More informationSoftware Datapath Acceleration for Stateless Packet Processing
June 22, 2010 Software Datapath Acceleration for Stateless Packet Processing FTF-NET-F0817 Ravi Malhotra Software Architect Reg. U.S. Pat. & Tm. Off. BeeKit, BeeStack, CoreNet, the Energy Efficient Solutions
More informationWhat s happened to the world of networking hardware offloads? Jesse Brandeburg Anjali Singhai Jain
What s happened to the world of networking hardware offloads? Jesse Brandeburg Anjali Singhai Jain 1 Agenda Introductions A Brief History of Offloads Hardware Offloads Future Look Proposals 2 Photo by
More informationMain Memory (Fig. 7.13) Main Memory
Main Memory (Fig. 7.13) CPU CPU CPU Cache Multiplexor Cache Cache Bus Bus Bus Memory Memory bank 0 Memory bank 1 Memory bank 2 Memory bank 3 Memory b. Wide memory organization c. Interleaved memory organization
More informationEvent Device Drivers. Release rc1
Event Device Drivers Release 19.02.0-rc1 December 23, 2018 CONTENTS 1 NXP DPAA Eventdev Driver 2 1.1 Features........................................ 2 1.2 Supported DPAA SoCs................................
More informationThe benefits and costs of writing a POSIX kernel in a high-level language
1 / 38 The benefits and costs of writing a POSIX kernel in a high-level language Cody Cutler, M. Frans Kaashoek, Robert T. Morris MIT CSAIL Should we use high-level languages to build OS kernels? 2 / 38
More informationAdvanced Computer Networks. End Host Optimization
Oriana Riva, Department of Computer Science ETH Zürich 263 3501 00 End Host Optimization Patrick Stuedi Spring Semester 2017 1 Today End-host optimizations: NUMA-aware networking Kernel-bypass Remote Direct
More informationUsing ebpf for network traffic analysis. Samuele Sabella Luca Deri
Using ebpf for network traffic analysis Samuele Sabella Luca Deri Network Monitoring In the past few years the need to better understand what s happening in our networks has lead from tools for volumetric
More informationKVM as The NFV Hypervisor
KVM as The NFV Hypervisor Jun Nakajima Contributors: Mesut Ergin, Yunhong Jiang, Krishna Murthy, James Tsai, Wei Wang, Huawei Xie, Yang Zhang 1 Legal Disclaimer INFORMATION IN THIS DOCUMENT IS PROVIDED
More informationLec 13: Linking and Memory. Kavita Bala CS 3410, Fall 2008 Computer Science Cornell University. Announcements
Lec 13: Linking and Memory Kavita Bala CS 3410, Fall 2008 Computer Science Cornell University PA 2 is out Due on Oct 22 nd Announcements Prelim Oct 23 rd, 7:30-9:30/10:00 All content up to Lecture on Oct
More informationAn interface for programmable IPv6 Segment Routing network functions in Linux
An interface for programmable IPv6 Segment Routing network functions in Linux Dissertation presented by Mathieu XHONNEUX for obtaining the Master s degree in Electrical Engineering Supervisor(s) Olivier
More informationHSA foundation! Advanced Topics on Heterogeneous System Architectures. Politecnico di Milano! Seminar Room A. Alario! 23 November, 2015!
Advanced Topics on Heterogeneous System Architectures HSA foundation! Politecnico di Milano! Seminar Room A. Alario! 23 November, 2015! Antonio R. Miele! Marco D. Santambrogio! Politecnico di Milano! 2
More informationLeveraging HyperTransport for a custom high-performance cluster network
Leveraging HyperTransport for a custom high-performance cluster network Mondrian Nüssle HTCE Symposium 2009 11.02.2009 Outline Background & Motivation Architecture Hardware Implementation Host Interface
More informationHDFI: Hardware-Assisted Data-flow Isolation
HDFI: Hardware-Assisted Data-flow Isolation Presented by Ben Schreiber Chengyu Song 1, Hyungon Moon 2, Monjur Alam 1, Insu Yun 1, Byoungyoung Lee 1, Taesoo Kim 1, Wenke Lee 1, Yunheung Paek 2 1 Georgia
More informationTesla GPU Computing A Revolution in High Performance Computing
Tesla GPU Computing A Revolution in High Performance Computing Gernot Ziegler, Developer Technology (Compute) (Material by Thomas Bradley) Agenda Tesla GPU Computing CUDA Fermi What is GPU Computing? Introduction
More informationTransparent Offloading and Mapping (TOM) Enabling Programmer-Transparent Near-Data Processing in GPU Systems Kevin Hsieh
Transparent Offloading and Mapping () Enabling Programmer-Transparent Near-Data Processing in GPU Systems Kevin Hsieh Eiman Ebrahimi, Gwangsun Kim, Niladrish Chatterjee, Mike O Connor, Nandita Vijaykumar,
More informationMemory: Page Table Structure. CSSE 332 Operating Systems Rose-Hulman Institute of Technology
Memory: Page Table Structure CSSE 332 Operating Systems Rose-Hulman Institute of Technology General address transla+on CPU virtual address data cache MMU Physical address Global memory Memory management
More informationProgrammable Dataplane
Programmable Dataplane THE NEXT STEP IN SDN? S I M O N J O U E T S I M O N. J O U E T @ G L A S G O W. A C. U K H T T P : / / N E T L A B. D C S.G L A. A C. U K GTS TECH+FUTURES WORKSHOP - SIMON JOUET
More informationMOVING FORWARD WITH FABRIC INTERFACES
14th ANNUAL WORKSHOP 2018 MOVING FORWARD WITH FABRIC INTERFACES Sean Hefty, OFIWG co-chair Intel Corporation April, 2018 USING THE PAST TO PREDICT THE FUTURE OFI Provider Infrastructure OFI API Exploration
More informationHSA Foundation! Advanced Topics on Heterogeneous System Architectures. Politecnico di Milano! Seminar Room (Bld 20)! 15 December, 2017!
Advanced Topics on Heterogeneous System Architectures HSA Foundation! Politecnico di Milano! Seminar Room (Bld 20)! 15 December, 2017! Antonio R. Miele! Marco D. Santambrogio! Politecnico di Milano! 2
More informationProgrammable NICs. Lecture 14, Computer Networks (198:552)
Programmable NICs Lecture 14, Computer Networks (198:552) Network Interface Cards (NICs) The physical interface between a machine and the wire Life of a transmitted packet Userspace application NIC Transport
More informationUse Dynamic Analysis Tools on Linux
Use Dynamic Analysis Tools on Linux FTF-SDS-F0407 Gene Fortanely Freescale Software Engineer Catalin Udma A P R. 2 0 1 4 Software Engineer, Digital Networking TM External Use Session Introduction This
More informationIsoStack Highly Efficient Network Processing on Dedicated Cores
IsoStack Highly Efficient Network Processing on Dedicated Cores Leah Shalev Eran Borovik, Julian Satran, Muli Ben-Yehuda Outline Motivation IsoStack architecture Prototype TCP/IP over 10GE on a single
More informationToward an ebpf-based clone of iptables
Toward an ebpf-based clone of iptables Matteo Bertrone, Sebastiano Miano, Jianwen Pi, Fulvio Risso, Massimo Tumolo Netdev 0x12, Montréal (Canada), July 12th, 2018 Objective Started in Nov 2017, with a
More informationHidden Linux Metrics with ebpf_exporter. Ivan Babrou
Hidden Linux Metrics with ebpf_exporter Ivan Babrou @ibobrik Performance team @Cloudflare What does Cloudflare do CDN Moving content physically closer to visitors with our CDN. Intelligent caching Unlimited
More informationSpring 2017 :: CSE 506. Device Programming. Nima Honarmand
Device Programming Nima Honarmand read/write interrupt read/write Spring 2017 :: CSE 506 Device Interface (Logical View) Device Interface Components: Device registers Device Memory DMA buffers Interrupt
More informationWORKLOAD CHARACTERIZATION OF INTERACTIVE CLOUD SERVICES BIG AND SMALL SERVER PLATFORMS
WORKLOAD CHARACTERIZATION OF INTERACTIVE CLOUD SERVICES ON BIG AND SMALL SERVER PLATFORMS Shuang Chen*, Shay Galon**, Christina Delimitrou*, Srilatha Manne**, and José Martínez* *Cornell University **Cavium
More informationNative Computing and Optimization. Hang Liu December 4 th, 2013
Native Computing and Optimization Hang Liu December 4 th, 2013 Overview Why run native? What is a native application? Building a native application Running a native application Setting affinity and pinning
More informationMoneta: A High-performance Storage Array Architecture for Nextgeneration, Micro 2010
Moneta: A High-performance Storage Array Architecture for Nextgeneration, Non-volatile Memories Micro 2010 NVM-based SSD NVMs are replacing spinning-disks Performance of disks has lagged NAND flash showed
More informationHIGH PACKET RATES IN GSTREAMER
HIGH PACKET RATES IN GSTREAMER GStreamer Conference 22 October 2017, Prague Tim-Philipp Müller < > tim@centricular.com MORE OF A CASE STUDY REALLY HIGH PACKET RATES? RTP RAW VIDEO: PACKET AND DATA RATES
More informationWriting high performance code. CS448h Nov. 3, 2015
Writing high performance code CS448h Nov. 3, 2015 Overview Is it slow? Where is it slow? How slow is it? Why is it slow? deciding when to optimize identifying bottlenecks estimating potential reasons for
More informationProcessors, Performance, and Profiling
Processors, Performance, and Profiling Architecture 101: 5-Stage Pipeline Fetch Decode Execute Memory Write-Back Registers PC FP ALU Memory Architecture 101 1. Fetch instruction from memory. 2. Decode
More informationThis Unit: Main Memory. Virtual Memory. Virtual Memory. Other Uses of Virtual Memory
This Unit: Virtual Application OS Compiler Firmware I/O Digital Circuits Gates & Transistors hierarchy review DRAM technology A few more transistors Organization: two level addressing Building a memory
More information/INFOMOV/ Optimization & Vectorization. J. Bikker - Sep-Nov Lecture 3: Caching (1) Welcome!
/INFOMOV/ Optimization & Vectorization J. Bikker - Sep-Nov 2015 - Lecture 3: Caching (1) Welcome! Today s Agenda: The Problem with Memory Cache Architectures Practical Assignment 1 INFOMOV Lecture 3 Caching
More informationBuffer Management for XFS in Linux. William J. Earl SGI
Buffer Management for XFS in Linux William J. Earl SGI XFS Requirements for a Buffer Cache Delayed allocation of disk space for cached writes supports high write performance Delayed allocation main memory
More informationSMP/SMT. Daniel Potts. University of New South Wales, Sydney, Australia and National ICT Australia. Home Index p.
SMP/SMT Daniel Potts danielp@cse.unsw.edu.au University of New South Wales, Sydney, Australia and National ICT Australia Home Index p. Overview Today s multi-processors Architecture New challenges Experience
More informationOverhead Evaluation about Kprobes and Djprobe (Direct Jump Probe)
Overhead Evaluation about Kprobes and Djprobe (Direct Jump Probe) Masami Hiramatsu Hitachi, Ltd., SDL Jul. 13. 25 1. Abstract To implement flight recorder system, the overhead
More information